# DIFFER: Disentangling Identity Features via Semantic Cues for Clothes-Changing Person Re-ID

arXiv: https://arxiv.org/abs/2503.22912 · HTML: https://mlsanity.com/paper/2503.22912/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 25 of its 43 references that are in the map.
- First posted on arXiv 28 Mar 2025.

## Builds on (most cited first, 25 of 25)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [Unsupervised Domain Adaptation by Backpropagation](https://mlsanity.com/paper/1409.7495/) (2014)
- [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://mlsanity.com/paper/2412.05271/) (2024)
- [EVA-CLIP: Improved Training Techniques for CLIP at Scale](https://mlsanity.com/paper/2303.15389/) (2023)
- [Beyond Part Models: Person Retrieval with Refined Part Pooling (and a Strong Convolutional Baseline)](https://mlsanity.com/paper/1711.09349/) (2017)
- [TransReID: Transformer-based Object Re-Identification](https://mlsanity.com/paper/2102.04378/) (2021)
- [CLIP-ReID: Exploiting Vision-Language Model for Image Re-Identification without Concrete Text Labels](https://mlsanity.com/paper/2211.13977/) (2022)
- [EVA-CLIP-18B: Scaling CLIP to 18 Billion Parameters](https://mlsanity.com/paper/2402.04252/) (2024)
- [Clothes-Changing Person Re-identification with RGB Modality Only](https://mlsanity.com/paper/2204.06890/) (2022)
- [Long-Term Cloth-Changing Person Re-identification](https://mlsanity.com/paper/2005.12633/) (2020)
- [Cloth-Changing Person Re-identification from A Single Image with Gait Prediction and Regularization](https://mlsanity.com/paper/2103.15537/) (2021)
- [Disentangling visual and written concepts in CLIP](https://mlsanity.com/paper/2206.07835/) (2022)
- [Learning Disentangled Representations via Mutual Information Estimation](https://mlsanity.com/paper/1912.03915/) (2019)
- [Improving Disentangled Text Representation Learning with Information-Theoretic Guidance](https://mlsanity.com/paper/2006.00693/) (2020)
- [Instruct-ReID: A Multi-purpose Person Re-identification Task with Instructions](https://mlsanity.com/paper/2306.07520/) (2023)
- [Learning Clothing and Pose Invariant 3D Shape Representation for Long-Term Person Re-Identification](https://mlsanity.com/paper/2308.10658/) (2023)
- [Semantic-aware Consistency Network for Cloth-changing Person Re-Identification](https://mlsanity.com/paper/2308.14113/) (2023)
- [Disentangling Representations of Text by Masking Transformers](https://mlsanity.com/paper/2104.07155/) (2021)
- [Disentangled Feature Representation for Few-shot Image Classification](https://mlsanity.com/paper/2109.12548/) (2021)
- [MLLMReID: Multimodal Large Language Model-based Person Re-identification](https://mlsanity.com/paper/2401.13201/) (2024)
- [ProFD: Prompt-Guided Feature Disentangling for Occluded Person Re-Identification](https://mlsanity.com/paper/2409.20081/) (2024)
- [Activity-Biometrics: Person Identification from Daily Activities](https://mlsanity.com/paper/2403.17360/) (2024)
- [Masked Attribute Description Embedding for Cloth-Changing Person Re-identification](https://mlsanity.com/paper/2401.05646/) (2024)

## Built on by (most cited first, 11 of 11)

- [Colors See Colors Ignore: Clothes Changing ReID with Color Disentanglement](https://mlsanity.com/paper/2507.07230/) (2025)
- [Try Harder: Hard Sample Generation and Learning for Clothes-Changing Person Re-ID](https://mlsanity.com/paper/2507.11119/) (2025)
- [LocalScore: Local Density-Aware Similarity Scoring for Biometrics](https://mlsanity.com/paper/2602.01012/) (2026)
- [DisenQ: Disentangling Q-Former for Activity-Biometrics](https://mlsanity.com/paper/2507.07262/) (2025)
- [The Devil Is in the Leakage: A Disentangled Dual-Purification Framework for High-Fidelity Hairstyle Transfer](https://mlsanity.com/paper/2607.11281/) (2026)
- [Not All Starting Points Are Equal: Pre-trained Priors and Their Outsized Impact on Person Identification](https://mlsanity.com/paper/2507.17640/) (2025)
- [CMCC-ReID: Cross-Modality Clothing-Change Person Re-Identification](https://mlsanity.com/paper/2604.02808/) (2026)
- [SapiensID 2.0: Aligning Human Recognition Foundation Models with Human Perception](https://mlsanity.com/paper/2608.10497/) (2026)
- [GaitCrafter: Diffusion Model for Biometric Preserving Gait Synthesis](https://mlsanity.com/paper/2508.13300/) (2025)
- [ICPR 2026 Competition on Privacy-Preserving Person Re-Identification from Top-View RGB-Depth Camera (TVRID)](https://mlsanity.com/paper/2605.04977/) (2026)
- [Learning Instance-Adaptive Low-Rank Orthogonal Subspaces for Clothes-Changing Person Re-Identification](https://mlsanity.com/paper/2606.11661/) (2026)

## Abstract (arXiv)

Clothes-changing person re-identification (CC-ReID) aims to recognize individuals under different clothing scenarios. Current CC-ReID approaches either concentrate on modeling body shape using additional modalities including silhouette, pose, and body mesh, potentially causing the model to overlook other critical biometric traits such as gender, age, and style, or they incorporate supervision through additional labels that the model tries to disregard or emphasize, such as clothing or personal attributes. However, these annotations are discrete in nature and do not capture comprehensive descriptions. In this work, we propose DIFFER: Disentangle Identity Features From Entangled Representations, a novel adversarial learning method that leverages textual descriptions to disentangle identity features. Recognizing that image features inherently mix inseparable information, DIFFER introduces NBDetach, a mechanism designed for feature disentanglement by leveraging the separable nature of text descriptions as supervision. It partitions the feature space into distinct subspaces and, through gradient reversal layers, effectively separates identity-related features from non-biometric features. We evaluate DIFFER on 4 different benchmark datasets (LTCC, PRCC, CelebreID-Light, and CCVID) to demonstrate its effectiveness and provide state-of-the-art performance across all the benchmarks. DIFFER consistently outperforms the baseline method, with improvements in top-1 accuracy of 3.6% on LTCC, 3.4% on PRCC, 2.5% on CelebReID-Light, and 1% on CCVID. Our code can be found here.